From 24ea591d2201c3257d666466e8fac50a6cf3c52f Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 6 Jul 2015 05:18:03 -0700 Subject: net: sched: extend percpu stats helpers qdisc_bstats_update_cpu() and other helpers were added to support percpu stats for qdisc. We want to add percpu stats for tc action, so this patch add common helpers. qdisc_bstats_update_cpu() is renamed to qdisc_bstats_cpu_update() qdisc_qstats_drop_cpu() is renamed to qdisc_qstats_cpu_drop() Signed-off-by: Eric Dumazet Cc: Alexei Starovoitov Acked-by: Jamal Hadi Salim Acked-by: John Fastabend Acked-by: Alexei Starovoitov Signed-off-by: David S. Miller --- net/core/dev.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'net/core/dev.c') diff --git a/net/core/dev.c b/net/core/dev.c index 6778a9999d52..e0d270143fc7 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -3646,7 +3646,7 @@ static inline struct sk_buff *handle_ing(struct sk_buff *skb, qdisc_skb_cb(skb)->pkt_len = skb->len; skb->tc_verd = SET_TC_AT(skb->tc_verd, AT_INGRESS); - qdisc_bstats_update_cpu(cl->q, skb); + qdisc_bstats_cpu_update(cl->q, skb); switch (tc_classify(skb, cl, &cl_res)) { case TC_ACT_OK: @@ -3654,7 +3654,7 @@ static inline struct sk_buff *handle_ing(struct sk_buff *skb, skb->tc_index = TC_H_MIN(cl_res.classid); break; case TC_ACT_SHOT: - qdisc_qstats_drop_cpu(cl->q); + qdisc_qstats_cpu_drop(cl->q); case TC_ACT_STOLEN: case TC_ACT_QUEUED: kfree_skb(skb); -- cgit From d746d707a8b1421a4ba46b497cb5d59e20161645 Mon Sep 17 00:00:00 2001 From: Anuradha Karuppiah Date: Tue, 14 Jul 2015 13:43:19 -0700 Subject: net core: Add protodown support. This patch introduces the proto_down flag that can be used by user space applications to notify switch drivers that errors have been detected on the device. The switch driver can react to protodown notification by doing a phys down on the associated switch port. Signed-off-by: Anuradha Karuppiah Signed-off-by: Andy Gospodarek Signed-off-by: Roopa Prabhu Signed-off-by: Wilson Kok Signed-off-by: David S. Miller --- net/core/dev.c | 20 ++++++++++++++++++++ 1 file changed, 20 insertions(+) (limited to 'net/core/dev.c') diff --git a/net/core/dev.c b/net/core/dev.c index 69445a33ace6..8810b6bbebfe 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -6074,6 +6074,26 @@ int dev_get_phys_port_name(struct net_device *dev, } EXPORT_SYMBOL(dev_get_phys_port_name); +/** + * dev_change_proto_down - update protocol port state information + * @dev: device + * @proto_down: new value + * + * This info can be used by switch drivers to set the phys state of the + * port. + */ +int dev_change_proto_down(struct net_device *dev, bool proto_down) +{ + const struct net_device_ops *ops = dev->netdev_ops; + + if (!ops->ndo_change_proto_down) + return -EOPNOTSUPP; + if (!netif_device_present(dev)) + return -ENODEV; + return ops->ndo_change_proto_down(dev, proto_down); +} +EXPORT_SYMBOL(dev_change_proto_down); + /** * dev_new_index - allocate an ifindex * @net: the applicable net namespace -- cgit From 0c4f691ff6791e55ac831666df0b49b1679c56e4 Mon Sep 17 00:00:00 2001 From: Scott Feldman Date: Sat, 18 Jul 2015 18:24:48 -0700 Subject: net: don't reforward packets already forwarded by offload device Just before queuing skb for xmit on port, check if skb has been marked by switchdev port driver as already fordwarded by device. If so, drop skb. A non-zero skb->offload_fwd_mark field is set by the switchdev port driver/device on ingress to indicate the skb has already been forwarded by the device to egress ports with matching dev->skb_mark. The switchdev port driver would assign a non-zero dev->offload_skb_mark for each device port netdev during registration, for example. Signed-off-by: Scott Feldman Acked-by: Jiri Pirko Acked-by: Roopa Prabhu Acked-by: Nicolas Dichtel Signed-off-by: David S. Miller --- net/core/dev.c | 10 ++++++++++ 1 file changed, 10 insertions(+) (limited to 'net/core/dev.c') diff --git a/net/core/dev.c b/net/core/dev.c index 8810b6bbebfe..2ee15afb412d 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -3061,6 +3061,16 @@ static int __dev_queue_xmit(struct sk_buff *skb, void *accel_priv) else skb_dst_force(skb); +#ifdef CONFIG_NET_SWITCHDEV + /* Don't forward if offload device already forwarded */ + if (skb->offload_fwd_mark && + skb->offload_fwd_mark == dev->offload_fwd_mark) { + consume_skb(skb); + rc = NET_XMIT_SUCCESS; + goto out; + } +#endif + txq = netdev_pick_tx(dev, skb, accel_priv); q = rcu_dereference_bh(txq->qdisc); -- cgit From f38a9eb1f77b296ff07e000823884a0f64d67b2a Mon Sep 17 00:00:00 2001 From: Thomas Graf Date: Tue, 21 Jul 2015 10:43:56 +0200 Subject: dst: Metadata destinations Introduces a new dst_metadata which enables to carry per packet metadata between forwarding and processing elements via the skb->dst pointer. The structure is set up to be a union. Thus, each separate type of metadata requires its own dst instance. If demand arises to carry multiple types of metadata concurrently, metadata dst entries can be made stackable. The metadata dst entry is refcnt'ed as expected for now but a non reference counted use is possible if the reference is forced before queueing the skb. In order to allow allocating dsts with variable length, the existing dst_alloc() is split into a dst_alloc() and dst_init() function. The existing dst_init() function to initialize the subsystem is being renamed to dst_subsys_init() to make it clear what is what. The check before ip_route_input() is changed to ignore metadata dsts and drop the dst inside the routing function thus allowing to interpret metadata in a later commit. Signed-off-by: Thomas Graf Signed-off-by: David S. Miller --- net/core/dev.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'net/core/dev.c') diff --git a/net/core/dev.c b/net/core/dev.c index 2ee15afb412d..cb52cba30ae8 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -7669,7 +7669,7 @@ static int __init net_dev_init(void) open_softirq(NET_RX_SOFTIRQ, net_rx_action); hotcpu_notifier(dev_cpu_callback, 0); - dst_init(); + dst_subsys_init(); rc = 0; out: return rc; -- cgit From b469139e81ca8265fb4797c007f8d3338f4191a5 Mon Sep 17 00:00:00 2001 From: "subashab@codeaurora.org" Date: Fri, 24 Jul 2015 03:03:29 +0000 Subject: dev: Spelling fix in comments Fix the following typo - unchainged -> unchanged Signed-off-by: Subash Abhinov Kasiviswanathan Signed-off-by: David S. Miller --- net/core/dev.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'net/core/dev.c') diff --git a/net/core/dev.c b/net/core/dev.c index cb52cba30ae8..4870c3556a5a 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -4995,7 +4995,7 @@ EXPORT_SYMBOL(netdev_all_upper_get_next_dev_rcu); * Gets the next netdev_adjacent->private from the dev's lower neighbour * list, starting from iter position. The caller must hold either hold the * RTNL lock or its own locking that guarantees that the neighbour lower - * list will remain unchainged. + * list will remain unchanged. */ void *netdev_lower_get_next_private(struct net_device *dev, struct list_head **iter) @@ -5050,7 +5050,7 @@ EXPORT_SYMBOL(netdev_lower_get_next_private_rcu); * Gets the next netdev_adjacent from the dev's lower neighbour * list, starting from iter position. The caller must hold RTNL lock or * its own locking that guarantees that the neighbour lower - * list will remain unchainged. + * list will remain unchanged. */ void *netdev_lower_get_next(struct net_device *dev, struct list_head **iter) { -- cgit From 906470c19da771e638e7c8e16e16c31995b139cc Mon Sep 17 00:00:00 2001 From: Phil Sutter Date: Tue, 18 Aug 2015 10:30:48 +0200 Subject: net: warn if drivers set tx_queue_len = 0 Due to the introduction of IFF_NO_QUEUE, there is a better way for drivers to indicate that no qdisc should be attached by default. Though, the old convention can't be dropped since ignoring that setting would break drivers still using it. Instead, add a warning so out-of-tree driver maintainers get a chance to adjust their code before we finally get rid of any special handling of tx_queue_len == 0. Signed-off-by: Phil Sutter Signed-off-by: David S. Miller --- net/core/dev.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'net/core/dev.c') diff --git a/net/core/dev.c b/net/core/dev.c index 4870c3556a5a..b1f3f4844e60 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -6997,6 +6997,9 @@ struct net_device *alloc_netdev_mqs(int sizeof_priv, const char *name, dev->priv_flags = IFF_XMIT_DST_RELEASE | IFF_XMIT_DST_RELEASE_PERM; setup(dev); + if (!dev->tx_queue_len) + printk(KERN_WARNING "%s uses DEPRECATED zero tx_queue_len - convert driver to use IFF_NO_QUEUE instead.\n", name); + dev->num_tx_queues = txqs; dev->real_num_tx_queues = txqs; if (netif_alloc_netdev_queues(dev)) -- cgit From 3b3ae880266d148bf73a573a766bc9b78c08d805 Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Wed, 26 Aug 2015 23:00:06 +0200 Subject: net: sched: consolidate tc_classify{,_compat} For classifiers getting invoked via tc_classify(), we always need an extra function call into tc_classify_compat(), as both are being exported as symbols and tc_classify() itself doesn't do much except handling of reclassifications when tp->classify() returned with TC_ACT_RECLASSIFY. CBQ and ATM are the only qdiscs that directly call into tc_classify_compat(), all others use tc_classify(). When tc actions are being configured out in the kernel, tc_classify() effectively does nothing besides delegating. We could spare this layer and consolidate both functions. pktgen on single CPU constantly pushing skbs directly into the netif_receive_skb() path with a dummy classifier on ingress qdisc attached, improves slightly from 22.3Mpps to 23.1Mpps. Signed-off-by: Daniel Borkmann Acked-by: Alexei Starovoitov Signed-off-by: David S. Miller --- net/core/dev.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'net/core/dev.c') diff --git a/net/core/dev.c b/net/core/dev.c index b1f3f4844e60..7bb24f1879b8 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -3657,7 +3657,7 @@ static inline struct sk_buff *handle_ing(struct sk_buff *skb, skb->tc_verd = SET_TC_AT(skb->tc_verd, AT_INGRESS); qdisc_bstats_cpu_update(cl->q, skb); - switch (tc_classify(skb, cl, &cl_res)) { + switch (tc_classify(skb, cl, &cl_res, false)) { case TC_ACT_OK: case TC_ACT_RECLASSIFY: skb->tc_index = TC_H_MIN(cl_res.classid); -- cgit From 0e4ead9d7b3655d76371604abb9b0dcc4e79bb7d Mon Sep 17 00:00:00 2001 From: Jiri Pirko Date: Thu, 27 Aug 2015 09:31:18 +0200 Subject: net: introduce change upper device notifier change info Add info that is passed along with NETDEV_CHANGEUPPER event. Signed-off-by: Jiri Pirko Signed-off-by: David S. Miller --- net/core/dev.c | 16 ++++++++++++++-- 1 file changed, 14 insertions(+), 2 deletions(-) (limited to 'net/core/dev.c') diff --git a/net/core/dev.c b/net/core/dev.c index 7bb24f1879b8..a8e6cf4298d3 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -5311,6 +5311,7 @@ static int __netdev_upper_dev_link(struct net_device *dev, struct net_device *upper_dev, bool master, void *private) { + struct netdev_notifier_changeupper_info changeupper_info; struct netdev_adjacent *i, *j, *to_i, *to_j; int ret = 0; @@ -5329,6 +5330,10 @@ static int __netdev_upper_dev_link(struct net_device *dev, if (master && netdev_master_upper_dev_get(dev)) return -EBUSY; + changeupper_info.upper_dev = upper_dev; + changeupper_info.master = master; + changeupper_info.linking = true; + ret = __netdev_adjacent_dev_link_neighbour(dev, upper_dev, private, master); if (ret) @@ -5367,7 +5372,8 @@ static int __netdev_upper_dev_link(struct net_device *dev, goto rollback_lower_mesh; } - call_netdevice_notifiers(NETDEV_CHANGEUPPER, dev); + call_netdevice_notifiers_info(NETDEV_CHANGEUPPER, dev, + &changeupper_info.info); return 0; rollback_lower_mesh: @@ -5462,9 +5468,14 @@ EXPORT_SYMBOL(netdev_master_upper_dev_link_private); void netdev_upper_dev_unlink(struct net_device *dev, struct net_device *upper_dev) { + struct netdev_notifier_changeupper_info changeupper_info; struct netdev_adjacent *i, *j; ASSERT_RTNL(); + changeupper_info.upper_dev = upper_dev; + changeupper_info.master = netdev_master_upper_dev_get(dev) == upper_dev; + changeupper_info.linking = false; + __netdev_adjacent_dev_unlink_neighbour(dev, upper_dev); /* Here is the tricky part. We must remove all dev's lower @@ -5484,7 +5495,8 @@ void netdev_upper_dev_unlink(struct net_device *dev, list_for_each_entry(i, &upper_dev->all_adj_list.upper, list) __netdev_adjacent_dev_unlink(dev, i->dev); - call_netdevice_notifiers(NETDEV_CHANGEUPPER, dev); + call_netdevice_notifiers_info(NETDEV_CHANGEUPPER, dev, + &changeupper_info.info); } EXPORT_SYMBOL(netdev_upper_dev_unlink); -- cgit From f84bb1eac0275283ccd76455e20f926e186ea8c8 Mon Sep 17 00:00:00 2001 From: Phil Sutter Date: Thu, 27 Aug 2015 21:21:36 +0200 Subject: net: fix IFF_NO_QUEUE for drivers using alloc_netdev Printing a warning in alloc_netdev_mqs() if tx_queue_len is zero and IFF_NO_QUEUE not set is not appropriate since drivers may use one of the alloc_netdev* macros instead of alloc_etherdev*, thereby not intentionally leaving tx_queue_len uninitialized. Instead check here if tx_queue_len is zero and set IFF_NO_QUEUE, so the value of tx_queue_len can be ignored in net/sched_generic.c. Fixes: 906470c ("net: warn if drivers set tx_queue_len = 0") Signed-off-by: Phil Sutter Signed-off-by: David S. Miller --- net/core/dev.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'net/core/dev.c') diff --git a/net/core/dev.c b/net/core/dev.c index a8e6cf4298d3..877c84834d81 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -7010,7 +7010,7 @@ struct net_device *alloc_netdev_mqs(int sizeof_priv, const char *name, setup(dev); if (!dev->tx_queue_len) - printk(KERN_WARNING "%s uses DEPRECATED zero tx_queue_len - convert driver to use IFF_NO_QUEUE instead.\n", name); + dev->priv_flags |= IFF_NO_QUEUE; dev->num_tx_queues = txqs; dev->real_num_tx_queues = txqs; -- cgit